A partial-trace matrix inequality and Werner-state distillability

Zhiwei Song, Lin Chen

GPT-5.5 Pro’s Proposition 2 expresses the rank-two partial-trace quadratic form q⁽²⁾₋₁/₂(C) as projection terms. Song and Chen then use a separate human-developed lemma to prove nonnegativity and two-copy nondistillability at the Werner threshold.

Sources for this record

What was already known

The two-copy threshold for NPT Werner states was a longstanding open problem, with earlier reductions to a rank-two partial-trace inequality and partial 4×4 progress. Several near-contemporaneous papers give independent solutions.

AI involvement AI-centered

Song and Chen credit GPT-5.5 Pro with the initial formulation and proof of Proposition 2; they checked it and supplied the remaining argument.

The model originated a retained central proposition, while the final inequality and Werner consequence also depend on human work.

Importance Notable

Provides a central step in one proof of the two-copy Werner-state threshold.

Limitations

The model did not supply the complete nondistillability proof. Fu, Gao and Park gave a separate solution, and other near-contemporaneous work reached the partial-trace result independently.

Technical verification Formal (human)

The authors explicitly report checking the model-generated Proposition 2 argument in full; this is human proof checking of the assessed AI contribution.

Formal checking

Checked the generated argument in full and presented its proof in Appendix A

Checker
Human authors of the partial-trace manuscript
Scope
Proposition 2 decomposition of q⁽²⁾₋₁/₂(C) for rank-at-most-two C
Basis
Reported by the linked checker
Evidence source

Checking a formal statement does not establish that it matches the original problem.

Human understanding Full

The authors explain how the projection decomposition leaves a single obstruction term and how their later lemma controls it.

Correction and publication history